Southeastern United States summer rainfall framework and its implication for seasonal prediction

摘要: A new rainfall framework is constructed to describe the complex probability distribution of southeastern United States (SE US) summer (June–July–August) rainfall, which cannot be well represented by traditional kernel fitting methods. The based on configuration a three-cluster finite normal mixture model and realized Bayesian inference Markov Chain Monte Carlo (MCMC) algorithm. three clusters reflect light, moderate, heavy in summer, are linked different climate factors. variation light intensity likely associated with combined effects La Nina tri-pole sea surface temperature anomaly (SSTA) over North Atlantic. Heavy concurs ‘horseshoe-like’ SSTA In contrast, moderate less correlated caused atmospheric internal dynamics. Rainfall characteristics their linkages SSTAs help improve seasonal predictions regional climate. Such has an important implication understanding response hydrology variability change; our study suggest that it can extended other regions seasons similar